Unleash the Power of AI with Keboola MCP Server and UBOS

In the rapidly evolving landscape of artificial intelligence, the ability to seamlessly connect AI agents with data sources and tools is paramount. The Keboola MCP (Model Context Protocol) Server, in conjunction with the UBOS platform, offers a robust solution to bridge this gap. This powerful combination allows you to expose your data, transformations, SQL queries, and job triggers to AI agents, eliminating the need for complex glue code and ensuring that the right data reaches the right agents, precisely when and where they need it.

What is MCP Server?

MCP (Model Context Protocol) is an open protocol that standardizes how applications provide context to LLMs. Think of it as a universal translator, enabling different AI models and tools to understand and interact with each other. The Keboola MCP Server acts as a vital intermediary, allowing AI models to access and interact with external data sources and tools within the Keboola ecosystem.

Key Benefits of Keboola MCP Server

• Seamless Integration: Connect your AI agents, MCP clients (like Cursor, Claude, Windsurf, and VS Code), and other AI assistants to Keboola without writing extensive custom code.
• Data Accessibility: Expose your valuable data assets, including tables, buckets, and metadata, to AI agents for analysis, insights, and decision-making.
• Transformation Capabilities: Empower AI agents to trigger and utilize SQL transformations within Keboola, enabling data manipulation and preparation for specific tasks.
• Job Orchestration: Allow AI agents to initiate and monitor Keboola jobs, automating data extraction, loading, and other critical processes.
• Enhanced Collaboration: Foster better communication and collaboration between AI models and your existing data infrastructure.
• Cost Efficiency: Reduce the time and resources required to integrate AI into your data workflows.

Use Cases for Keboola MCP Server

The Keboola MCP Server unlocks a wide range of use cases for AI-powered data interaction:

• AI-Driven Data Analysis: Enable AI agents to analyze your data directly within Keboola, uncovering hidden patterns and generating actionable insights.

  Example: An AI agent could analyze sales data by region to identify underperforming areas and suggest targeted marketing campaigns.

• Automated Data Pipeline Management: Automate the creation, execution, and monitoring of data pipelines using AI agents.

  Example: An AI agent could create a SQL transformation to join customer and order tables, then trigger a data extraction job to update a reporting dashboard.

• Natural Language Querying of Data: Allow users to query your Keboola data using natural language, making data access more intuitive and accessible.

  Example: A user could ask, “What are the top 10 customers by revenue?” and receive a direct answer from the AI agent.

• AI-Powered Documentation and Metadata Management: Use AI agents to automatically generate and maintain documentation for your Keboola projects.

  Example: An AI agent could search Keboola documentation based on natural language queries or update table descriptions based on data analysis.

• AI Agent orchestration Orchestrate AI Agent to follow your workflow process. You can set it up to the process with a single end-point. You can define different scenarios and make AI agent follow your requirements.

Key Features of Keboola MCP Server

The Keboola MCP Server boasts a rich set of features designed to facilitate seamless AI integration:

• Storage Management:
  • Query tables directly using AI agents.
  • Manage table and bucket descriptions through natural language commands.
  • Retrieve detailed information about buckets and tables.
• Component Management:
  • List and inspect extractors, writers, data apps, and transformation configurations.
  • Retrieve detailed configuration information for specific components.
• SQL Transformation:
  • Create SQL transformations using natural language, simplifying data manipulation.
  • Identify the SQL dialect (Snowflake or BigQuery) used in your workspace.
• Job Management:
  • Run components and transformations directly from AI agents.
  • Retrieve detailed job execution details.
  • List and filter jobs by status, component, or configuration.
• Metadata Management:
  • Search, read, and update project documentation and object metadata using natural language queries.

Getting Started with Keboola MCP Server

To begin using the Keboola MCP Server, you will need:

• Python 3.10+ installed on your system.
• Access to a Keboola project with administrative rights.
• Your preferred MCP client (e.g., Claude, Cursor).

Configuration Steps

1. Obtain Credentials: Gather the following credentials from your Keboola project:
   • KBC_STORAGE_TOKEN: Your Keboola Storage API token (refer to the Keboola documentation for instructions on creating and managing tokens).
   • KBC_WORKSPACE_SCHEMA: Your workspace schema (refer to the Keboola documentation for instructions on obtaining your workspace schema).
   • Keboola Region: Specify Keboola Region.
2. Configure MCP Client: Configure your chosen MCP client (Claude or Cursor) to connect to the Keboola MCP Server. This typically involves providing the server’s command, arguments, and environment variables.
3. Run the MCP Server: Choose one of the available methods for running the MCP Server:
   • Integrated Mode (Recommended): Allow your MCP client to automatically start the server.
   • Local Development Mode: Run the server from your local development environment.
   • Manual CLI Mode: Run the server manually in a terminal (primarily for testing and debugging).
   • Docker: Deploy the server using Docker.

Configuration Examples

Claude Desktop Configuration

{ “mcpServers”: { “keboola”: { “command”: “uvx”, “args”: [ “keboola_mcp_server”, “–api-url”, “https://connection.YOUR_REGION.keboola.com” ], “env”: { “KBC_STORAGE_TOKEN”: “your_keboola_storage_token”, “KBC_WORKSPACE_SCHEMA”: “your_workspace_schema” } } } }

Cursor Configuration

{ “mcpServers”: { “keboola”: { “command”: “uvx”, “args”: [ “keboola_mcp_server”, “–api-url”, “https://connection.YOUR_REGION.keboola.com” ], “env”: { “KBC_STORAGE_TOKEN”: “your_keboola_storage_token”, “KBC_WORKSPACE_SCHEMA”: “your_workspace_schema” } } } }

Verify Your Setup

Once the MCP Server is configured and running, verify the connection by sending a simple query to your Keboola project:

What buckets and tables are in my Keboola project?

Why UBOS for AI Agent Development?

UBOS is a full-stack AI Agent Development Platform that empowers businesses to build and deploy custom AI agents across various departments. By integrating with the Keboola MCP Server, UBOS provides a comprehensive solution for connecting AI agents with your enterprise data.

UBOS Key Features

• AI Agent Orchestration: Design and manage complex AI agent workflows with ease.
• Enterprise Data Connectivity: Seamlessly connect AI agents with your existing data sources.
• Custom AI Agent Development: Build custom AI agents using your preferred LLM model.
• Multi-Agent Systems: Create sophisticated AI systems that leverage multiple agents to achieve complex goals.
• Scalability and Reliability: Deploy and scale AI agents with confidence on the UBOS platform.

UBOS Integration with Keboola MCP Server

The UBOS platform seamlessly integrates with the Keboola MCP Server, providing a unified environment for building, deploying, and managing AI agents that interact with your Keboola data. This integration allows you to:

• Develop AI agents within UBOS that can directly query and analyze your Keboola data using the MCP Server.
• Automate data pipeline management tasks within Keboola using AI agents orchestrated by UBOS.
• Create custom AI-powered applications that leverage the data and transformation capabilities of Keboola.

Conclusion

The Keboola MCP Server, combined with the UBOS platform, offers a powerful solution for unlocking the full potential of AI in your organization. By seamlessly connecting AI agents with your data assets and tools, this integration empowers you to automate tasks, gain deeper insights, and drive better business outcomes. Embrace the future of AI-powered data interaction with Keboola MCP Server and UBOS.